We study visual attention because of its importance in visual perception. Behavioral paradigms, such as change blindness, have shown us that while we think we perceive the whole visual world, we only take away information about regions or objects that we have attended. Because visual attention is a foundation of visual perception, it underlies most of our interactions with the perceived world - both our physical interactions and more intellectual interactions, such as learning and memory. This is highlighted in patients with parietal lesions - the difficulty they have in interacting with the world leads to significantly loger hospital stays than many other stroke patients, incurring greater costs for family members and the community as a whole. Thus, increasing our understanding of the mechanisms underlying the guidance of attention is important as a health issue in addition to being critical in allowing s to gain a deeper insight into how the brain makes decisions based on both external and cognitive inputs and, in the long run, insight into the mechanisms underlying visual perception itself. In the previous grant, we focused on the role of the lateral intraparietal area (LIP), and showed how the activity in LIP can be used to guide attention. The results from that study suggested that the information in LIP needs to be processed in a specific way before it can be used to guide eye movements. In this study, we build on this work by testing predictions based on the hypothesis that that there is a reciprocal connection from the frontal eye field (FEF) to LIP, in which a subset of neurons in FEF provides top-down input to LIP and a subset of neurons in FEF receives processed information from LIP. We do so by examining the responses of neurons in FEF and identifying them as projecting to LIP or receiving projections from LIP. In aim 1, we test the prediction that FEF neurons encode signals which could be the genesis of top-down responses seen in LIP. We will show that neurons that project directly or indirectly back to LIP have early responses which contain component signals that are adequate and appropriate to explain top-down effects in LIP. These responses often begin before an eye movement is made and are often overridden by a feed-forward connection 150-200 ms after the eye stops moving. In aim 2, we test the prediction that later responses in FEF neurons are consistent with the normalized output of responses seen in LIP. We do this by comparing the responses in FEF to the behavior and estimates of stimulus value. Based on the responses seen in earlier cortical areas, we have explicit predictions about how neurons that receive this information should behave. Overall, these experiments are aimed at showing how the LIP-FEF circuit guides the allocation of attention and, in doing so, will give us a greater understanding o how neural circuits process information in general sensori-motor transformations.